自动相关监视目标的跟踪算法研究

Tracking Algorithm of Automatic Dependent Surveillance Target

为了实现对新航行系统自动相关监视目标的跟踪,通过对机动目标跟踪算法进行分析研究,给出了一种基于高斯和的贝叶斯估计多模型跟踪算法。其主要思想是对多个模型滤波器滤波输出进行加权平均,每个模型滤波器采用推广卡尔曼滤波算法,当需要进行状态预测时,应用卡尔曼滤波理论进行K步迭代递推估计。该算法是解决非线性、非高斯估计问题的一个较好的变结构多模型算法。基于高斯和的贝叶斯估计算法对匀速、匀加速和强机动目标都可以达到良好的跟踪效果。

In order to track automatic dependent surveillance （ADS） targets of Future Air Navigation systems （FANS） ,a multiple-model （MM） tracking algorithm based on Bayesian estimation using Gaussian sums is given after several target tracking algorithms are analysed.The main idea of the algorithm is to get the weighted average of the output of different filters. Each of the filters adopts the Extended Kalman Filtering algorithm, and uses Kalman filtering theory for iterative estimation by K steps when states of target are to be predicted. It is a good structure-variable MM algorithm for resolving problems of nonlinear and nongaussian estimation. Algorithm based on Bayesian estimation using Gaussian sums may obtain good tracking result for targets with constant velocity, constant acceleration and high maneuverability.